Obfuscating Gender in Social Media Writing

By Sravana Reddy, Kevin Knight. 2016

100 citations recorded by the source corpus.

The vast availability of textual data on social media has led to an interest in algorithms to predict user attributes such as gender based on the user’s writing. These methods are valuable for social science research as well as targeted advertising and proﬁling, but also compromise the privacy of users who may not realize that their personal idiolects can give away their demographic identities. Can we automatically modify a text so that the author is clas-siﬁed as a certain target gender, under limited knowledge of the classiﬁer, while preserving the text’s ﬂuency and meaning? We present a basic model to modify a text using lexical substitution, show empirical results with Twitter and Yelp data, and outline ideas for extensions.

Subfields: Artificial Intelligence, Information Systems

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Information Systems

Citations in the corpus

Mitigating Gender Bias in Natural Language Processing: Literature Review (2019, background): “Other important aspects such as model/data trans-parency (Mitchell et al., 2019; Bender and Fried-man, 2018) and privacy preservation (Reddy and Knight, 2016; Elazar and Goldberg, 2018; Li et al., 2018) are also not covered in this literature survey.”
Deceiving Google's Perspective API Built for Detecting Toxic Comments (2017, methodology): “In a similar work, the authors proposed a method for gender obfuscating in social media writing [22].”
Politeness Transfer: A Tag and Generate Approach (2020, background): “, 2018), gender and political slant transfer (Reddy and Knight, 2016; Prabhumoye et al., 2018), and formality transfer (Rao and Tetreault, 2018; Xu et al.”
Style Transfer Through Back-Translation (2018, methodology): “tions, deletions, and insertions to modulate demographic properties of a writer (Reddy and Knight, 2016), integrating stylistic and demographic speaker traits in statistical machine translation (Rabinovich et al.”
“Transforming” Delete, Retrieve, Generate Approach for Controlled Text Style Transfer (2019, methodology): “…from adapting conversational style in dialogue agents (Zhou et al., 2017), obfus-cating personal attributes (such as gender) to prevent privacy intrusion (Reddy and Knight, 2016), altering texts to be more formal or informal (Rao and Tetreault, 2018), to generating poetry (Yang et al., 2018).”
Gradient-based Constrained Sampling from Language Models (2022, background): “They also find useful applications in anonymizing protected attributes in written text (Reddy and Knight, 2016), as writing assistants to avoid users’ implicit biases (Ma et al., 2020; Field and Tsvetkov, 2020”
Detect and Perturb: Neutral Rewriting of Biased and Sensitive Text via Gradient-based Decoding (2021, unknown): “Example 3 is a review by a female (from Yelp dataset (Reddy and Knight, 2016)) while Example 4 describes a female applicant in a graduate admission reference letter (our data).”
Judging a Book by its Description : Analyzing Gender Stereotypes in the Man Bookers Prize Winning Fiction (2018, background): “A related task is to modify or paraphrase text data to obfuscate gender as in (Reddy and Knight, 2016) Another closely related work is to change the style of the text to different levels of formality as in (Rao and Tetreault, 2018).”
Towards Robust Toxic Content Classification (2019, background): “Reddy and Knight (2016) showed that the gender of posters on social media can be obfuscated by using a background corpus to identify words indicative of each gender and replacing those words with semantically similar words.”
Adhering, Steering, and Queering: Treatment of Gender in Natural Language Generation (2020, methodology): “In current work, this is achieved by removing and substituting related words that are associated with stereotypical gendered language [53,57] (such as removing emotional language, as frequently associated with female authors - as described in the Background, above).”
Controlled Text Generation as Continuous Optimization with Multiple Constraints (2021, background): “They also have many other positive use-cases, for example, anonymizing personal attributes in written text [54], and even aiding authors in avoiding implicit biases in their writing [44, 14].”
Text Style Transfer: A Review and Experiment Evaluation (2020, methodology): “The researchers constructed the gender dataset 9 by first preprocessing a Yelp review dataset annotated with the gender of the reviewers [64] by split the reviews into sentences and preserve the gender label for each sentence.”
Attacking Automatic Video Analysis Algorithms: A Case Study of Google Cloud Video Intelligence API (2017, methodology): “s or generating out-of-distribution samples [ 9,19 ,29 ,32 ,35 ,47 ,56 ]. This property has been used to attack voice interfaces [ 15 ], face-recognition systems [ 40 ] and text classication systems [23, 39]. Attacks on ML systems can be also classied according to the adversary'saccesstothesyst”
Fair Transfer of Multiple Style Attributes in Text (2019, background): “One of the early work by [13] performs a lexicon based substitution on the words of the source sentence in order to transform its author style from male to female.”
Multiple-Attribute Text Rewriting (2018, methodology): “…English according to a fastText (Joulin et al., 2016) classifier, 2) not about restaurants, 3) rated 3/5 stars as they tend to be neutral in sentiment (following Shen et al. (2017)), or 4) where the gender is not identifiable by the same method as in Reddy & Knight (2016); Prabhumoye et al. (2018).”
It’s going to be okay: Measuring Access to Support in Online Communities (2018, unknown): “We hope to also use our method to better support privacy-preserving behavior (Allen and Wiles, 2016; Reddy and Knight, 2016) by help-”
Anonymized BERT: An Augmentation Approach to the Gendered Pronoun Resolution Challenge (2019, background): “Gender bias has been an important topic in natural language processing in recent years (Bolukbasi et al., 2016; Reddy and Knight, 2016; Chiappa and Gillam, 2018; Madaan et al., 2018).”